Factors affecting medical artificial intelligence (AI) readiness among medical students: taking stock and looking forward

Background

Measuring artificial intelligence (AI) readiness among medical students is essential to assess how prepared future doctors are to work with AI technology. Therefore, this study aimed to examine the factors influencing AI readiness among medical students at Kermanshah University of Medical Sciences, both by evaluating the current situation and considering future developments.

Methods

This was a cross-sectional descriptive-analytical study. The statistical population consisted of 800 first- to fifth-year medical students selected through convenient sampling at Kermanshah University of Medical Sciences from November to March 2023. The data collection tools were demographic checklists and Persian version questionnaire of the medical artificial intelligence readiness scale for medical students (MAIRS-MS). The data were analyzed at a significance level of P < 0.05 using independent t-test, and analysis of variance (ANOVA) tests through SPSS-24 software.

Results

Most of the students were male (56.13%). The overall score for medical AI readiness was 70.59 ± 19.24 out of a maximum possible score of 110. Students had the highest mean score of 9.73 ± 2.96 out of 15 in vision and the lowest mean score of 25.74 ± 7.52 out of 40 in ability. The overall mean of AI readiness (71.84 ± 18.27) was higher in females than males (69.62 ± 19.93), but this difference was not significant (p = 0.106). Furthermore, the mean total score of AI readiness increased with the increasing age of the students.

Conclusion

Our findings underscore the need to prepare students to work with AI technologies and to provide them with the essential knowledge and skills across different areas of AI. Accordingly, the Kermanshah University of Medical Sciences student’s education unit should set up more AI training centers to provide and introduce basic artificial intelligence courses. Moreover, universities should identify the needs of students based on scientific evidence, and the medical education system should design AI training programs in its educational framework in the same direction.

Artificial intelligence (AI), which can be traced back to the 1950s when artificial neural networks were developed [1], is a remarkable scientific breakthrough that has changed the modern world and is advancing rapidly [2]. AI generally refers to the machine’s ability to imitate human intelligent behavior [3]. Van Der and Bleakley and Topol [4, 5] suggested that AI can make a significant contribution to more reliable diagnosis, improved treatment outcomes, and reduced medical malpractice. Nowadays, AI has attracted a lot of attention in medicine due to its many applications in daily life activities [6]. Additionally, it has grown significantly in healthcare and medicine [7]. Current developments in health and medicine indicate that AI technologies will soon be used by many doctors in various fields [5]. Due to its global growth, AI is expected to be considered one of the main components of medical education [7]. AI system helps doctors in various fields, including early diagnosis of many infectious diseases, cardiovascular diseases [8], neurological diseases such as Parkinson’s [9], diseases of the digestive system [10], electrocardiogram [11], caries diagnosis [12], colorectal polyps [13], and prostate cancer [14]. Therefore, AI professional training is an inevitable necessity [15].

According to previous studies [7, 16,17,18,19], medical students have a positive attitude towards artificial intelligence and are interested in incorporating it into their educational system and curriculum, emphasizing the necessity and application of AI for medical students. According to previous studies, doctors will use AI more objectively in the future and will understand it better [20]. Doctors will perform many of their tasks by AI in the coming years, and healthcare services will be provided at a faster rate [20]. Thus, new learning needs should be considered to change the professional identity of doctors in such a way that doctors can acquire the necessary skills to participate in the new technology of the future and AI enrich students’ perception of clinical problems [21]. Improvements should be made in the curricula of medical students for them to better understand the components of AI to get the most benefit from AI technologies [22, 23].

Curriculum planners should plan the content of the courses based on the needs of the learners after knowing the readiness of the students concerning the main concepts [19]. The learning process in medical education and healthcare is expected to change in the future with the growing trend of AI [24]. The perceived AI readiness of medical students is important for curriculum development and needs analysis. Despite the importance of AI, there will be challenges until its place in the medical education system is understood [17, 25].

One of the challenges facing medical education is to optimally include AI in curricula to support the new roles of doctors in the future [4, 26]. The readiness of students is important before AI is included in the medical education system. It is useful to measure students’ readiness with the perspective of where education should begin [27]. The guidelines are given to the students according to their characteristics, their needs are identified, and educational programs are developed accordingly by specifying the students’ readiness.

Therefore, it’s really important for medical students to have opportunities, both in their courses and extracurricular activities, to learn about the clinical applications and ethical perspectives of artificial intelligence tools [28]. However, the current teaching approaches in medical schools are limited in covering AI in the curriculum [29]. Nowadays, medical students have an optimistic attitude towards the concept of AI and are keen to see it included in their studies [16]. These students believe that in the near future, AI will have a profound impact on medical education [30]. However, as it stands, most medical schools do not include AI in their curricula [31].

Medical schools must quickly adapt to these conditions to equip the next generation of physicians with the skills needed to adopt AI in their future work due to the rapid development of AI in healthcare [32]. Although AI has many applications to improve the clinical management of patients, it is still unclear to what extent medical students are prepared and use it in their education and later in medical practice [33]. In this way, measuring AI readiness among medical students to determine the readiness of future doctors to work with AI technology is an inevitable necessity [34, 35]. Although digital health has brought enormous changes in medicine, the contribution of many concepts such as “AI readiness” to teaching and learning in medical education is not well defined, and significant scientific evidence is not available in this field. Given the importance of the subject and the lack of extensive research in this area, which highlights a gap in informational resources, this study has been conducted with the aim of determining the level of AI readiness and identifying the factors influencing AI readiness among medical students at Kermanshah University of Medical Sciences.

Study design

This was a cross-sectional descriptive-analytical study. The statistical population consisted of 800 first- to fifth-year medical students selected through convenient sampling at Kermanshah University of Medical Sciences from November to March 2023. The inclusion criteria were studying at a university during the study period, consent to participate in the study, and being a first- to fifth-year medical student. The exclusion criterion was incomplete answering of the scales.

The data collection tools include

The demographic information of students (gender, age, marital status, native status, and medical school year);

Medical AI Readiness Scale for Medical Students (MAIRS-MS), which was developed by Karaca et al. to assess medical AI readiness among medical students [36]. This scale has 22 questions and four scales. Cognition measures students’ cognition of AI terms and logic, as well as data science (questions 1–8). Ability measures students’ readiness to select the appropriate AI application, appropriate use of the application, and their ability to explain the application to patients (questions 9–16). Vision measures students’ ability to explain the limitations, strengths, and weaknesses of AI in medicine and predict future opportunities and threats (questions 19 − 17). Ethics evaluates students’ ability to adhere to legal and ethical regulations when using AI technologies (questions 20–22). The answers are scored using a 5-point Likert scale (1 = completely disagree to 5 = completely agree). Cronbach’s alpha values ​​were 0.83 for cognition, 0.77 for ability, 0.72 for vision, and 0.63 for ethics. The validity and reliability of the Iranian scale developed by Moodi Ghalibaf et al., [35]among students at Mashhad Medical School were as follows: 0.88 for cognition, 0.90 for ability, 0.86 for vision, and 0.85 for ethics. Cronbach’s alpha value for the whole scale was 0.94. In a study conducted by Rezazadeh et al., [34]on medical students in Kerman, Cronbach’s alpha coefficient for the whole scale was 0.94. In the present study, the reliability coefficient of the scale for cognition, ability, vision, and ethics were 0.91, 0.92, 0.89, and 0.92, respectively, with a total Cronbach’s alpha coefficient of 0.966. The final mean score of all respondents in each area was calculated to assess their AI readiness.

Data collection

After obtaining the necessary permits from the Research Vice-Chancellor of Kermanshah University of Medical Sciences, obtaining the code of ethics under the number IR.KUMS.REC.1403.192, and submitting it to the Faculty of Medical Sciences, the required permits for the study were obtained. The researchers (AZ and MAA) explained the objectives of the study to the participants after introducing themselves and obtaining the consent of the research units to participate in the study. They were included in the study if they met all the inclusion criteria and obtained written and informed consent to participate in the study. According to the explanations provided at the beginning of the scale, participation in the study was free and students could freely not fill the scale at any time and leave the study. A Microsoft Forms based online survey questionnaire (prepared in persin language) was sent to all the participants via Telegram communication groups. The data obtained from the scales were recorded confidentially without including the participant’s name and information and were only used for the study. The participants sent all responses by entering the electronic survey form.

Data analysis

After confirming, completing, and collecting the scales, it was time to extract and process the data. For this purpose, the questions were coded and directed. We performed all analyses on complete case data. The data were analyzed at a significance level of P < 0.05 using the independent t-test, and analysis of variance (ANOVA) tests through SPSS-24 software (Inc., Chicago, Ill., USA). Descriptive statistics including frequency, standard deviation, mean, and percentage were used to describe the demographic characteristics of the participants.

351 (43.88%) of the 800 first- to fifth-year medical students who participated in the study were female and 449 (56.13%) were male. According to the findings, the samples were in the age range of 17 to 30 years. 735 people (91.88%) were single and 65 people (8.13%) were married. 510 people (63.75%) were natives of Kermanshah province. 288 people (36%) lived in dormitories, 67 people (8.38%) had rented houses, and 445 people (55.63%) had private houses. 197 people were studying medicine in the first year, 157 people in the second year, 176 people in the third year, 143 people in the fourth year, and 127 people in the fifth year. The samples were divided almost equally in each academic year from the first to the fifth year. (Table 1). According to the leveling of AI readiness among medical students in cognition, ability, vision, and ethics, higher scores with a mean of 9.73 ± 2.96 and 9.15 ± 3.36 out of 15 belonged to vision and ethics, respectively. The lower scores with a mean of 25.98 ± 7.36 and 25.74 ± 7.52 out of 40 belonged to cognition and ability, respectively. The total medical AI readiness among medical students was 70.59 ± 19.24 out of 110 (Table 2). Students at Kermanshah University of Medical Sciences had the worst performance in the ability domain and the best performance in the vision domain.

The mean AI readiness among male and female students was compared using T-test. The overall mean of AI readiness (71.84 ± 18.27) was higher in females than males (69.62 ± 19.93), but this difference was not significant (p = 0.106). Furthermore, the mean total score of AI readiness increased with the increasing age of the students.

The students aged less than 20 years had the lowest mean (69.55 ± 18.87), and those in the age group of 26 to 30 years had the highest mean score of AI readiness (72.33 ± 16.78), but this difference was not statistically significant. Non-native students had a higher mean AI readiness (71.17 ± 19.17) than native students (70.26 ± 19.29), but this difference was not statistically significant (p = 0.516). Married students had a higher mean AI readiness (72.26 ± 18.42) than unmarried students (70.44 ± 19.32), but this difference was not statistically significant (p = 0.466) (Table 3). The results of AI readiness based on 4 MAIRS-MS scale domains including cognition, ability, vision, and ethics can be seen in Table 4. According to the findings, the respondents evaluated their readiness as lower in ability and higher in cognition, vision, and ethics. This can be seen in their mean MAIRS-MS scores, as provided in (Table 3).

This study was conducted to investigate the factors affecting medical AI readiness among medical students of Kermanshah University of Medical Sciences with taking stock and looking forward. The medical artificial intelligence readiness scale for medical students (MAIRS-MS) including cognition, ability, vision, and ethics was measured through the mean total scores of the students. The results showed that students reported their readiness higher in the areas of ethics and vision. This was in line with the results of a study by Xuan et al., [37]. In a study by Aboalshamat et al., AI readiness levels were assessed among medical and dental students and graduates in Saudi Arabia through the MAIRS-MS. The results suggested that the participants had low AI readiness levels, and it was recommended that AI be taught from the very beginning of the medical education years [38].

Another study in Jordan showed that students scored higher in the ability domain with a mean of 22.57. The study found that academic performance, as measured by Grade Point Average (GPA), was positively associated with overall AI readiness (P = 0.023). Additionally, prior exposure to AI through formal education or practical experience significantly enhanced readiness (P = 0.009). In contrast, AI readiness levels did not show significant variation across different medical schools in Jordan [39].

Xuan et al., measured AI readiness among medical students in Malaysia using MAIRS-MS. Policymakers are recommended to launch more AI training courses to introduce the basics. The concepts of AI, especially for general medical students, enable them to gain more confidence in interacting with AI technology in the future [37].

Although there are few studies on students’ readiness levels, other studies on vision indicate positive results in this field. For example, according to Park et al., students had a positive vision of the role of medical education in the future of medicine [40]. Bisdas et al. argued that medical school students were interested in AI and had an optimistic vision of its impact on medical education [16]. Having a positive vision of an issue seems to indicate its acceptance. Another study similarly showed that students who have perceived AI are ready to face it and are more receptive to working with it [17]. Also, mean total scores the results suggested a lower level of students’ readiness in cognition and ability. This was not consistent with the results of the study by Xuan et al., [37]. They argued that medical AI readiness among medical students was sufficient, with students reporting high readiness levels in all areas of cognition, ability, vision, and ethics. Most students were very interested in AI and were optimistic about its applications in medicine [37]. One of the reasons for the low AI readiness in the areas of cognition and ability among students of the University of Medical Sciences seems to be the high speed of entering AI and digital health in medical education without a proper foundation in that society. Students have not yet developed the necessary cognitive adaptation, or they may not have achieved a correct assessment of their cognition and ability in this field [41].

The results of t test showed that female students were more ready than male students, but this difference was not statistically significant. However, according to the results of a study by Sarwar et al., male respondents were more confident in using AI applications and were less afraid of AI technologies, indicating higher readiness of male students than female students [42].

Based on the results obtained from ANOVA analysis, AI readiness increased as students aged, with students in the 26–30 age group reporting the highest mean total AI readiness scores. However, this difference was not statistically significant. According to Xuan et al., there was a significant relationship between age and positive vision of AI. A positive vision leads to more AI readiness [37]. It seems that increasing cognitive growth in adulthood can play an important role in this field. Personal factors and beliefs can affect a person’s perception of new issues such as AI with increasing age and cognitive development [43].

Although AI has brought about significant changes in all fields, especially medical science, and the American Medical Association (AMA) promoted the implementation of AI and raised interesting discussions about issues concerning payment, regulation, deployment, and liability in its 2019 guidelines [32], AI seems to have received less attention in Iran’s medical education systems, and accurate scientific evidence of students’ AI readiness in different populations is not available. However, a scientific report suggests that students strongly believe that AI will profoundly affect medical education in the future, indicating the need to pay attention to all its aspects [30].

ANOVA analysis showed that there was no significant relationship between students’ years of study and AI readiness. However, Xuan et al. argued that preclinical students have a higher degree of ability, vision, and ethics compared to clinical students, but there was no significant relationship between years of study and cognition [37]. The results of a survey of 321 medical students in Ontario found that the majority of students (79%) argued that their medical education did not adequately make them ready to work with AI tools or applications. They agreed that more training in the medical program was needed to increase their level of AI readiness [44]. Iranian politicians seem to have failed to make room for AI in the medical education system. The lack of students’ knowledge of AI and its applications in medicine due to the lack of a proper educational setting is one of the issues that has received less attention in Iran’s educational system and students are not aware of its necessity.

Medical education requires serious changes in the educational structure due to changes in the healthcare system due to AI. Artificial intelligence has no place in the curriculum of medical schools today [31]. Curriculum improvements should be made in medical school programs to better understand the basic components and algorithms of AI to achieve maximum efficiency of artificial intelligence-based technologies in medical procedures and protect the medical professional identity [22, 23]. For example, medical students who have been trained in AI have been reported to feel more secure in working with AI in the future than students who have not been trained [7]. Another study suggests that prior AI training becomes a critical skill for clinicians to interpret the medical literature, evaluate potential clinical software developments, and formulate research questions [45]. In this way, AI is expected to transform the learning process in medical education and healthcare services by moving into a new era [37].

Although the use of AI can be very helpful, the noteworthy point is that it should not only replace the roles of doctors and professors in today’s universities but may also change the roles [33]. Due to these developments, the current knowledge of medical students and practicing doctors in AI is likely to strengthen such an evolution [46, 47].

Recommendations

Future qualitative studies are recommended to investigate the attitudes and perceptions of medical students toward AI in various populations in such a way that universities can identify the needs of students based on the scientific evidence of their population and the medical education system can design AI training programs in its educational framework accordingly. Future studies are also recommended to consider potential changes in medical AI readiness over time to improve medical AI knowledge and skills among medical students to ensure a healthy AI ecosystem that leads to the development of innovative AI technologies and companies.

In this study, the students of the University of Medical Sciences reported their level of readiness in cognition and ability to be low. Accordingly, the educational unit of the university is recommended to set up more AI training centers to introduce basic courses in this field. More AI courses should be offered to younger populations to engage with AI digital information earlier and gain more confidence in interacting with AI technology in the future. Students should be encouraged to engage with and expose themselves to learning about AI technology. The more students are exposed to AI technology, the higher their level of AI readiness will be and will ensure greater confidence to collaborate with AI technologies in medicine in the future. Policymakers and medical educators should set up more AI training centers to provide online or offline AI training courses. Medical schools should include medical AI in their programs in a more interesting way and provide appropriate practical AI training for medical students.

Strengths and limitations

This study represents the first large-scale investigation in western Iran to assess the readiness of medical students for artificial intelligence. Additionally, the data collected for this research was complete with no missing variables. However, the interpretation of the findings must be considered within the context of several limitations. First, selection bias may have influenced the results due to the use of convenience sampling. Second, the study’s limited generalizability must be noted, as it only included students from Kermanshah University of Medical Sciences. Caution is needed when attempting to apply the findings to other populations. Moreover, since the research was conducted at a public university, the results may not extend to private institutions. Another limitation concerns the data collection process. The use of Telegram to gather questionnaire responses may have introduced several challenges. Participants might have faced technical difficulties, such as connectivity issues or app-specific glitches, which could have hindered their ability to respond effectively. Furthermore, there were concerns about the security of data shared on these platforms, potentially affecting confidentiality and data protection. Finally, not all individuals in the target population may have been active on these platforms, which could have led to sample bias.

The datasets generated and/or analysed during the current study are not publicly available due to consent not being obtained from participants for this purpose but are available from the corresponding author on reasonable request.

MAIRS-MS:

Medical artificial intelligence readiness scale for medical students

AI:

Artificial intelligence

UME:

Undergraduate medical education

ML:

Machine learning

MD:

Doctor of medicine

The research team appreciates all the participants for providing their valuable knowledge and experiences. This study is the result of research project No. 4030310 approved by the Student Research Committee of Kermanshah University of Medical Sciences.

This study did not receive any funding from funding agencies.

Authors and Affiliations

1. Cardiovascular Research Center, Health Policy and Promotion Institute, Kermanshah University of Medical Sciences, Kermanshah, Iran

   Arash Ziapour, Parisa Janjani & Mohammad Amin Amani

2. Psychology Research Centre, Khazar University, Baku, Azerbaijan

   Arash Ziapour & Murat Yıldırım

3. Department of Public Health, Asadabad School of Medical Sciences, Asadabad, Iran

   Fatemeh Darabi

4. Department of Psychology, Faculty of Science and Letters, Agri Ibrahim Cecen University, Ağrı, Türkiye, Turkey

   Murat Yıldırım

5. Student Research Committee, Kermanshah University of Medical Sciences, Kermanshah, Iran

   Sayeh Motevaseli

Contributions

All authors were responsible for the study. AZ and SM conceived and designed the survey. AZ and PJ performed the investigation. SM, AZ and MAA analyzed the data. AZ, FD and MY revised the paper. MY edited the paper grammatically. All the authors have read and approved the final manuscript.

Corresponding author

Correspondence to Sayeh Motevaseli.

Ethics approval and consent to participate

The data collection in the present study was conducted after the approval of Kermanshah University of Medical Sciences, and Publication Ethics Boards the number IR.KUMS.REC.1403.192. We confirm that all methods used in this study were carried out in accordance with relevant guidelines and regulations. The participation of students was completely voluntary and informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions